Polychlorinated Biphenyls in Major Foodstuffs on the Canadian Market

23 

Full text

(1)

KE030G Självständigt arbete för kandidatexamen i kemi,

30 högskolepoäng

Polychlorinated Biphenyls in Major

Foodstuffs on the Canadian Market

Kjell Hope

Orebro University MTM Research Centre

June 24, 2018

Supervisor: Dr. Heidelore Fiedler Examiner: Dr. Tuulia Hyötilainen

(2)

Table of Contents

1 Introduction ... 3

2 Materials and Methods ... 5

2.1 Chemicals ... 5 2.2 Equipment ... 5 2.3 Standards ... 5 2.4 Samples ... 6 2.5 Method ... 7 2.5.1 Extraction ... 7 2.5.2 Instrumentation ... 8

3 Results and Discussion ... 11

4 Conclusion ... 20

(3)

1 Introduction

Polychlorinated biphenyls (PCBs) are synthetic chlorinated hydrocarbon compounds that consist of two benzene rings linked by a single carbon-carbon bond. Each of the ten hydrogen atoms can be replaced with chlorine atoms. This creates the potential for 209 specific congeners to exist. PCBs have been produced commercially since 1929(IARC, 2016). Their major use has been as dielectric or hydraulic fluids in capacitors and condensers but they have also been used in open applications such as plasticizers, surface coatings, inks, adhesives, flame retardants, pesticide extenders, paints, and microencapsulation of dyes for carbonless duplicating paper. Because PCBs are resistant to both acids and bases and are relatively heat-stable, they do not break down in sun-light and are resistant to biological transformation. PCB bioaccumulate in wildlife and in humans and undergo long-range transport. This led to global regulation and PCBs were among the initial persistent organic pollutants (POPs) in the Stockholm Convention on Persistent Organic Pollutants (UNEP, 2001).

All congeners of PCBs are lipophilic and their lipophilicity increases with increasing degree of chlorination. Due to this they exert very little water solubility. Congeners with a lower degree of chlorination are more volatile than those with a higher degree, caused by a lower molecular weight. Like many organochlorine compounds, most of the congeners are highly persistent and accumulate within the food chain. Investigations in many parts of the world have revealed widespread distribution of PCBs in the environment (Holoubek, 2001). Further environmental contamination may occur from the disposal of old electrical equipment containing PCBs. Many countries have severely restricted or banned the production of PCBs. The more highly

chlorinated PCB congeners adsorb strongly to soil and sediment and are not found in water at significant concentrations. The various congeners in soil and sediment have half-lives that extend from months to years. Adsorption of PCBs generally increases with the extent of chlorination of the congener. Volatilization and biodegradation are two very slow processes and are the major pathways of PCB removal from water and soil. Some of the desirable properties of PCBs are what make them so persistent in the environment. PCBs are commonly measured as part of biomonitoring studies as their lipophilic properties allow them to bioaccumulate in the food chain (Drouillard et al., 2013).

Consumption of PCB contaminated foods is the most common route of exposure for the general human population (Fitzgerald et al., 1996). All PCB are toxic and were classified as human carcinogens – Group 1 - by the International Agency on Research for Cancer (IARC, 2016). Many but not all of the effects of PCBs are through interactions with the arylhydrocarbon receptor (AhR) (Bemis et al. , 2005). The ability for different PCB congeners to bind to the AhR is results in the fact to exhibit “dioxin-like” effects. These effects are the same as from 2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD). Therefore, these PCBs are included into the scheme of the toxic equivalents and have been assigned toxic equivalency factors (TEFs) (van den Berg et al., 2006). These 12 dioxin-like PCBs are as follows: PCB 77, 81, 105, 114, 118, 123, 126, 156, 157, 167, 169, 189. Another group of PCBs are the so-called “indicator PCBs”: six congeners (PCB 28, 52, 101, 138, 153, 180) are found at high concentrations in the technical mixtures as

described above and also in the environment, food, and in human tissues. These congeners are do not exhibit dioxin-like toxicity but are toxic through other mechanisms.. They are also Group 1 carcinogens according to IArC classification. Due to their higher concentration, indicator PCBs

(4)

are used to indicate contamination from PCBs. The PCB 118 congener is considered both a marker and dioxin-like PCB

Most of what is known about the human health effects of PCBs is based on exposures due to accidental releases or work-related activities. These exposures are much higher than those for the general population through environmental or dietary exposure. The adverse health effects include a severe form of acne (chloracne), swelling of the upper eyelids, discolouring of the nails and skin, numbness in the arms and/or legs, weakness, muscle spasms, chronic bronchitis, and problems related to the nervous system. In addition, the International Agency for Research on Cancer (IARC) has concluded that there is sufficient evidence to link long-term, high-level PCBs exposure in occupational settings to an increased incidence of cancer, particularly liver and kidney cancer (IARC, 2016).

Studies performed in Europe indicate that the main foods responsible for current PCB exposure are fatty fish, meat products, and dairy products (Adamse, 2016). One study performed in Belgium used a food basket approach to discern dietary intake of PCBs from the products

mentioned above. The results showed that fish had the highest average concentration for the sum of PCBs of 7.1 ng/g wet weight. Conclusions from a Finnish study supports that fish have the highest PCB concentration with a maximum concentration of 2.5 ng/g wet weight. The TEQ observed for the fish samples was 1.5 pg TEQ/g (Kiviranta, 2004). Both examinations of PCBs in food products show that there are measurable levels present. Very little research has been done on the potential PCB exposure to Canadians through what they eat. There are little to no

regulations regarding these compounds in Canada (Canadian Environmental Protection Act, 1999). Existing information focussed on more populous centres in eastern Canada such as Toronto and Ottawa. This is mainly due to the work that has been done concerning the Great Lakes and how they are a sink for pollutants due to the heavy industrialization in the area. In contrast, Western Canada is a centre for agriculture in North America. This is due in part to the climate in the region which, unlike the rest of Canada, has milder winters and summers. Larger areas of land are present which creates the ability for the growth of livestock such as chicken and cows. Another major food source in the region stems from the aquatic life. British Columbia is home to a vast array of salmon and other fresh water fish such as trout. The wild salmon return to spawn in many rivers across the province. These fish have been swimming in the Pacific Ocean and are susceptible to the contaminants present in their food source. Regulations concerning the availability fishing of wild salmon has led to the rise of fish farms. These salmon are unable to live freely in the ocean and are kept in small lots their entire lives. Their only contact with the outside world is through the feed that is given to them by the farmers. The aim of this study is to determine where there are significant concentrations of PCBs in the Canadian market, and if so, what regulation should be put into place to protect the health of the Canadian population.

(5)

2 Materials and Methods

2.1 Chemicals

• 3N hydrochloric acid • Concentrated Sulfuric Acid • Toluene • 9:1 Toluene:methanol • 1:1 Toluene:hexane • Hexane • Acetone • Dichloromethane (DCM) • Isooctane • Diatomaceous Earth

• Acid silica gel column (15 mm and 25 mm columns) (CAPE Technologies, City, Country)

• Carbon column (CAPE Technologies, City, Country)

2.2 Equipment

• 250 ml beaker • Scoopula • Aluminum foil • 60 ml test tubes • 15 ml test tubes

• 1 L amber widemouth jar

• Top Drive Standard Roller apparatus • Glass? Plastic?? Pipettes and pipette bulbs • Hand-held blender

• Nitrogen concentrator apparatus

• CAPE Technologies tandem carbon/acid silica column apparatus • 750 ul GC vial

• RocketTM Evaporator System (company, city, country)

• Dionex™ ASE™ Dionium extraction cells and filter (company, city, country) • Dionex™ ASE™ accelerated solvent extraction system (company, city, country) • TSQ 9000 triple quadrupole GC-MS/MS (Thermo ScientificTM, city, country)

• Thermo ScientificTM DFSTM Magnetic Sector GC-HRMS (Thermo Scientific,

city, country)

2.3 Standards

• PCB WHO and Marker Internal Standard

o P48-M-ES. 13C marker PCBs (Wellington Laboratories, Guelph, Canada) o P48-W-ES. 13C WHO PCBs (Wellington Laboratories, Guelph, Canada)

(6)

• PCB Spike Standard

o BP-MS. PCB congener solution (Wellington Laboratories, Guelph, Canada)

o BP-MS2. Chlorobiphenyl solution/mix (Wellington Laboratories, Guelph, Canada)

• PCB Recovery Standard

o P48-RS. 13C PCBs recovery standard (Wellington Laboratories, Guelph, Canada)

• PCDD/PCDF Internal Standard

o EPA-1613LCS. 13C PCDD & PCDFs (Wellington Laboratories, Guelph, Canada)

• PCDD/PCDF Spike Standard

o EDF-5493. 20 compound mix (Cambridge Isotope Laboratories,

Tewksbury, MA, USA) • PCDD/PCDF Recovery Standard

o EPA-1613ISSS. 13C – 1234-TCDD, 13C – 123789-HxCDD (Wellington Laboratories, Guelph, Canada)

2.4 Samples

• Ground beef

o Grandville Island o Fresh Street Market o Save On Foods • Eggs (12)

o Canadian Harvest o Avalon

o Maple Hill Farms • Milk

o Dairyland (3.5%) o Avalon (3.5%)

o The Farm House (4.0%) • Salmon

o Fresh Street Market Sockeye (Wild) o Grandville Island Spring (Farm) o Safeway Atlantic (Farm)

• Trout

o Save On Foods Steelhead (Farm) o Fresh Street Market Steelhead (Wild) o Grandville Island Rainbow (Wild)

(7)

2.5 Methods

2.5.1 Extraction

During the collection process, samples were kept in a regulated fridge at a temperature of 4 0C.

Once all samples were collected, they were homogenized based on their specific matrices. Fish samples were de-skinned and if necessary, de-boned so that only the fillet would be used for analysis. Homogenization was done using a hand blender. Eggs and milk were also mixed using a hand blender. Ground meat samples were assumed to already be a homogenous mixture. The analytical procedure was based on EPA 1668C and adjusted in order to follow in house methods. Depending on the matrix, approximately 10 g of fish, 10 g of egg, 25 g of beef, and 200 grams of milk were taken for extraction and each sample was extracted in duplicate. Samples were mixed with 10 g of diatomaceous earth and placed in 100 ml dionium ASE cells. Carbon 13 labeled congeners similar to the analytes of interest were used as internal standards to determine the efficiency of the extraction method. Samples had one to five ng of PCB and 500 to 1000 pg of PCDD/PCDF internal standard added. Blanks were used to evaluate contamination resulting from the analytical procedure. Laboratory control samples (LCS) were used to determine if the method produced false negatives. The recoveries of the spiked analytes were evaluated for extraction accuracy. LCS contained one ng of native PCB and two hundred to one thousand picogram of native PCDD/PCDF. Both blanks and LCS were run with each batch to ensure quality of extraction.

Samples were extracted using accelerated solvent extraction (ASE) using 90:10 (v/v)

toluene/methanol. Three cycles were performed with five-minutes hold times. Once extraction was complete, samples were poured into pre-weighed 60 ml VOCs and concentrated using a Rocket Evaporator to dryness. Here lipid determination was calculated. Samples were re-constituted in 10 ml of hexane and eluted through 25 mm acid silica clean up columns. As each column can only handle a maximum of 2.5 g of fat, depending on the amount of lipid present, more than one column was used per sample.

Milk samples were extracted using 3N hydrochloric for acid digestion. The same amount of standards were added as mentioned above. These were placed on a tumbler over the weekend (3 days). After, 50 ml of hexane was added to the samples and tumbled over night (minimum of 16 h). This extract was collected. Two more aliquots of hexane were added to the samples and tumbled for four and one hour, respectively. Similar to the ASE extraction, samples were concentrated to dryness for lipid determination and eluted through and acid silica column for lipid removal.

All extracts were mixed with 10 ml of concentrated sulfuric acid and left to sit overnight (16 h). Further clean up was done using CAPE Technologies pressurized tandem acid silica/carbon column. Carbon columns were washed using 10 ml aliquots of toluene, DCM, and hexane under a constant stream of nitrogen while the acid silica had 30 ml of hexane eluted by gravity. Once cleaned and primed, the carbon column was inserted onto the bottom of the silica column. Sample was added and eluted under pressure with 30 ml of hexane for the first fraction making sure to not let the carbon column go dry. The carbon column was removed and attached to an empty column where 6 ml of 1:1 toluene:hexane was eluted for fraction 2 again making sure to

(8)

not let the carbon column go dry. Finally, the carbon column was flipped upside down and re attached to the empty column where 50 ml of toluene eluted for fraction 3. Fractions 1 and 2 were collected together and represent the PCB portion of the sample where fraction 3 contained all the PCDD/PCDF.

The PCB fraction was concentrated to 100 µl in DCM. In order to determine instrumentation efficiency 2 ng of PCB recovery standard added prior to instrumental analysis.

PCDD/PCDF fraction was concentrated to 20 µl in isooctane and had 1 ng of PCDD/PCDF recovery standard added prior to instrumentation.

Calibration verifications were run at the beginning and end of each sequence as well as ever 12 hours of instrumental operation to ensure data would be in agreement with the calibration. Each sample was extracted in duplicate to confirm that the results obtained were reproducible.

2.5.2 Instrumentation

A new method was developed to run all 209 congener PCBs on a triple quadrupole instrument. This was thought to be appropriate as there are several benefits in using this instrument as

opposed to the traditional method with uses a magnetic sector high resolution mass spectrometer. The triple quadrupole is a much more stable and rugged instrument, as well as it exhibits only a minor sensitivity loss when compared to the HRMS. It is more selective for the different congener groups with less column bleed through to the tetrachlorobiphenyls as well as

pentachlorobiphenyls which is caused by a loss of 2 chlorine atoms from the higher substituted groups. This is facilitated as the MS/MS monitors daughter ions using different collision energies which were optimized to produce the best chromatographic results. The MS/MS requires

significantly less maintenance such as source cleaning which increases the uptime of the instrument and also makes it much more cost effective over the HRMS. It is also easier to train new technicians to use the instrument.

PCB samples were run on a triple quadrupole GC-MS/MS Thermo ScientificTM TSQ 9000 using a 1 µl split/splitless injector with conditions shown below (Table 1). A 5 µl Programable

Temperature Vaporizing (PTV) injector was also used. This injector can be rapidly heated and cooled allowing the sample to be injected in liquid form and then vaporized to remove the solvent so only the analytes are introduced to the column.

Table 1: Instrumental parameters for the TSQ 900 triple quadrupole GC-MS/MS for PCB analysis

TSQ

Inlet Method PTV Method

Mode Splitless with Surge PTV mode Splitless

Base temperature (0C) 230 Temperature (0C) 40.0

Constant flow (mL/min) 1.4 Injection pressure (kPa) 70.0

Oven Method Injection time (min) 0.03

(9)

Two new SGE HT8 column (60 m x 0.25 mm i.d.) sent by the manufacturer were used to test the new method on the MS/MS. These were sent to test new unknown linings that the manufacturer had installed in the columns. After running calibration standards as well as a 209-congener native standard, there was no negligible difference in selectivity between the two columns. It was also possible to have baseline resolution between PCB 127 and 105.

Collision energy optimization was performed in the method development. It was found that modifying these conditions significantly effected the lower chlorinated congeners consisting of three or less chlorine atoms where as the heavier congeners showed little change.

PCBs were also run using a Magnetic Sector GC-HRMS Thermo ScientificTM DFSTM with 2 µl injection volume. A SGE HT8 column (60 m x 0.25 mm i.d.) was used for all instrumentation. PCDD/PCDF samples ran on a Magnetic Sector GC-HRMS Thermo ScientificTM DFSTM with 1 µl injection volume. A RTX Dioxin-2 (60 m x 0.25 mm i.d, 0.25 um film thickness)

Table 2: Instrumental parameters for GC-HRMS for analysis of PCDD/F and PCBs

HRMS

Inlet Method (PCDD/F) Inlet Method (PCB)

Mode Mode Splitless

Temperature (0C) 280 Temperature (0C) 220

Constant flow (mL/min) 1.4 Constant flow (mL/min) 1.5

Oven Method Oven Method

Initial temperature (0C) 90.0 Initial temperature (0C) 40.0

Initial time (min) 2.0 Initial time (min) 1.0

Ramp #1 (0C/min) 40.0 Ramp #1 (0C/min) 40.0

Final temp. #1 (0C) 248 Final temp. #1 (0C) 185

Hold Time #1 (min) 0 Hold Time #1 (min) 0

Ramp #2 (0C/min) 0.2 Ramp #2 (0C/min) 1.5

Final temp. #2 (0C) 252 Final temp. #2 (0C) 240

Hold time #2 (min) 0 Hold time #2 (min) 0

Initial time (min) 1.0 Evaporation rate (0C/s) 14.5

Ramp #1 (0C/min) 40.0 Evaporation time (min) 0.06

Final temp. #1 (0C) 155 Transfer pressure (kPa) 210

Hold Time #1 (min) 0.5 Transfer rate (0C/s) 14.5

Ramp #2 (0C/min) 2.0 Transfer temp. (0C) 240

Final temp. #2 (0C) 290 Transfer time (min) 0.9

Hold time #2 (min) 0.0 Cleaning rate (0C/s) 14.5

Ramp #3 (0C/min) 17.0

Cleaning temp. (0C) 280

Final temp. #3 (0C) 350 Cleaning time (min) 1.0

(10)

Ramp #3 (0C/min) 5.0 Ramp #3 (0C)/min) 4.0

Final temp. #3 (0C) 265 Final temp. #3 (0C) 305

Hold time #3 (min) 3.0 Hold time #3 (min) 3.0

Ramp #4 (0C/min) 2.0

Final temp. #4 (0C) 285.0

Hold time #4 (min) 0

Ramp #5 (0C/min) 15.0

Final temp. #5 (0C) 339

Hold time #5 (min) 13.0

Isotopic dilution was used for both HRMS and GC-MS/MS to identify and quantify data. Results were generated using Target QuanTM 3.1.

(11)

3 Results and Discussion

While sampling, special consideration was taken to obtain a variety of different matrices as well as diverse samples within each sample group using a general food basket approach with specific focus on products similar to those chosen in European studies (Voorspoels, 2008). The major difference between the current study and those done in Europe is the sampling size. Those studies are much more comprehensive with samples numbering several hundred whereas a small batch of 15 samples were used in this study. The choice of supermarkets was based off of

varying customers each grocery store/market attracts in order to get a general overview of what products could be purchased. Grandville Island is a very small market that focuses on the

freshness of it products. Fresh Street Market advertises their products as sustainable and organic. Safeway and Save On Foods caters to the average shopper who focuses more on the price point of the purchases.

The variety of fish samples were chosen to exhibit both differing species as well as aquatic environment. Trout was used to observe the fresh water concentrations where as the salmon focused on salt water. Furthermore, in farmed fish the concentrations are influenced by the housing (aquarium) and the feed that the fish are fed whereas the wild fish are exposed to their natural environment (their habitat).

Multiple runs were performed using different injectors and detection instruments in order to compare the results using different techniques. Time permitting, further tests could have been run to confirm the results seen such as performing re-extraction of samples.

A formal method detection limit study was not run on the TSQ 9000 to confirm the LODs. Instead, the values were taken from the lowest concentration standard run on the instrument. Further method development would be needed to decrease these values in order to compete with the GC-HRMS system.

Table 3 shows the summary of the limit of detection (LOD) and limits of quantification (LOQ) for the different combinations,Different values were found when comparing the LOD and LOQs exhibited between the two instruments . Having lower values gives greater reliability that the detected analyte is actually present at low concentrations and can be distinguished from a blank. Majority of the marker PCBs found greatly exceeded the detection limits on either instrument. However, the dioxin-like PCBs were not generally seen with the exception of PCB105 and PCB118 with values ranging from 10 to 730 pg/g.

(12)

Table 3: Calculation of the congener specific PCB LOD and LOQ for the TSQ and HRMS. LOD is calculated using 3σ. LOQ is 10σ.

TSQ (1 ul injection) HRMS Markers Lowest standard

pg/g LOD pg/g LOQ pg/g Stdev pg/g LOD pg/g LOQ pg/g PCB 28 0.500 1.5 5 0.174 0.522 1.74 PCB 52 0.500 1.5 5 0.150 0.45 1.5 PCB 101 2.500 7.5 25 0.141 0.423 1.41 PCB 138 0.500 1.5 5 0.158 0.474 1.58 PCB 153 0.500 1.5 5 0.107 0.321 1.07 PCB 180 0.500 1.5 5 0.157 0.471 1.57 dl-PCB PCB 77 0.500 1.5 5 0.090 0.27 0.9 PCB 81 2.000 6 20 0.129 0.387 1.29 PCB 126 0.500 1.5 5 0.108 0.324 1.08 PCB 169 2.000 6 20 0.113 0.339 1.13 PCB 105 0.500 1.5 5 0.140 0.42 1.4 PCB 114 0.100 0.3 1 0.130 0.39 1.3 PCB 118 0.500 1.5 5 0.140 0.42 1.4 PCB 123 0.500 1.5 5 0.126 0.378 1.26 PCB 156 0.500 1.5 5 0.085 0.255 0.85 PCB 157 2.000 6 20 0.120 0.36 1.2 PCB 167 2.000 6 20 0.133 0.399 1.33 PCB 189 2.000 6 20 0.100 0.3 1

Tables 4 and 5 as well as Appendix 1 depicts PCB concentrations found in the samples,

separating dl-PCBs and the markers. Upper bound as well as lower bound TEQs were calculated to show the possible variation in concentrations among the different instrumental methods. Fish samples were calculated based on wet weight (ww) where all terrestrial samples used lipid weight (lw) Table 6 also shows the upper and lower bound concentrations for the PCDD/PCDF analysis.

(13)

Table 4: Upper bound and lower bound WHO2005 TEQ concentrations of dioxin-like PCBs obtained by three different instrumentation methods. Lower bound values were calculated by substituting 0 for all non-detects. Upper bound values had the respective LOQs substituted in for non-detects.

TSQ 1 ul Fish-SH

farm

Fish-SH wild Fish-Rainbow Fish-Salmon PAC wild Fish-Salmon PAC farm Fish-Salmon Atlantic Milk-Dairyland Milk-Avalon Milk-Fhouse Egg-CDN Harvest Egg-Avalon Egg-MHF Ground beef-Regular Ground beef-Lean fresh Ground beef-Superma rket regular Lower bound Unit pg/g ww pg/g lipid

WHO2005-TEQno-PCB 0.000 0.000 0.049 0.000 0.802 0.000 0.000 0.000 0.105 0.000 0.000 0.000 0.000 0.000 0.000

WHO2005-TEQmo-PCB 0.005 0.003 0.007 0.005 0.032 0.003 0.000 0.003 0.001 0.001 0.000 0.001 0.001 0.001 0.001

WHO2005-TEQPCB 0.005 0.003 0.056 0.005 0.833 0.003 0.000 0.003 0.106 0.001 0.000 0.001 0.001 0.001 0.001

Σdl-PCB 173.556 97.738 235.915 166.356 1071.142 115.897 0.000 95.462 42.833 26.966 8.447 26.012 28.745 30.530 22.960

WHO2005-TEQtotal 0.005 0.003 0.056 0.005 0.833 0.003 0.000 0.003 0.106 0.001 0.000 0.001 0.001 0.001 0.001

Upper bound

WHO2005-TEQno-PCB 1.107 1.107 0.555 1.101 1.402 1.107 1.107 1.107 0.711 1.107 1.107 1.107 1.107 1.107 1.107

WHO2005-TEQmo-PCB 0.007 0.004 0.007 0.006 0.032 0.005 0.002 0.004 0.003 0.003 0.003 0.003 0.003 0.003 0.003

WHO2005-TEQPCB 1.113 1.111 0.563 1.107 1.434 1.111 1.109 1.111 0.715 1.110 1.109 1.109 1.109 1.110 1.109

Σdl-PCB 269.556 193.738 271.489 242.356 1096.142 211.897 131.000 191.462 158.833 152.966 134.447 147.012 144.745 151.530 143.960

WHO2005-TEQtotal 1.113 1.111 0.563 1.107 1.434 1.111 1.109 1.111 0.715 1.110 1.109 1.109 1.109 1.110 1.109

TSQ PTV 5 ul

Lower bound

Unit pg/g ww pg/g ww pg/g ww pg/g ww pg/g ww pg/g ww pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid

WHO2005-TEQno-PCB 0.014 0.294 0.074 0.018 0.411 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

WHO2005-TEQmo-PCB 0.005 0.005 0.007 0.005 0.031 0.004 0.005 0.003 0.001 0.001 0.001 0.001 0.000 0.000 0.001

WHO2005-TEQPCB 0.019 0.299 0.081 0.023 0.442 0.004 0.006 0.003 0.001 0.001 0.001 0.001 0.000 0.000 0.001

Σdl-PCB 164.325 167.743 247.535 167.858 1057.475 124.894 171.004 95.462 45.012 45.054 25.020 42.461 0.000 0.000 29.836

WHO2005-TEQtotal 0.019 0.299 0.081 0.023 0.442 0.004 0.006 0.003 0.001 0.001 0.001 0.001 0.000 0.000 0.001

Upper bound

WHO2005-TEQno-PCB 0.514 0.894 0.574 0.618 0.411 1.107 1.107 1.107 0.707 1.107 1.107 1.107 1.107 1.107 1.107

WHO2005-TEQmo-PCB 0.005 0.005 0.007 0.006 0.031 0.004 0.006 0.004 0.003 0.003 0.003 0.003 0.002 0.002 0.002

WHO2005-TEQPCB 0.519 0.899 0.581 0.624 0.442 1.110 1.114 1.111 0.710 1.110 1.109 1.110 1.109 1.109 1.108

Σdl-PCB 175.325 187.743 258.535 212.827 1057.475 182.094 262.004 191.462 156.012 166.054 146.020 163.461 131.000 131.000 110.83

6

(14)

HRMS 2 ul

Lower bound

Unit pg/g ww pg/g ww pg/g ww pg/g ww pg/g ww pg/g ww pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid pg/g lipid

WHO2005-TEQno-PCB 0.000 0.000 0.000 0.144 0.001 0.000 0.000 0.000 0.000 0.076 0.000 0.000 0.000 0.000 0.000

WHO2005-TEQmo-PCB 0.003 0.002 0.005 0.003 0.022 0.003 0.004 0.003 0.001 0.001 0.000 0.001 0.001 0.001 0.001

WHO2005-TEQPCB 0.003 0.002 0.006 0.147 0.023 0.003 0.004 0.003 0.001 0.077 0.000 0.001 0.001 0.001 0.001

Σdl-PCB 95.462 78.399 181.038 102.528 733.784 88.877 149.522 95.462 39.814 24.331 14.052 20.037 35.227 27.970 22.265

WHO2005-TEQtotal 0.003 0.002 0.006 0.147 0.023 0.003 0.004 0.003 0.001 0.077 0.000 0.001 0.001 0.001 0.001

Upper bound

WHO2005-TEQno-PCB 0.142 0.142 0.142 0.144 0.143 0.143 0.142 0.142 0.142 0.110 0.142 0.142 0.142 0.142 0.142

WHO2005-TEQmo-PCB 0.003 0.002 0.005 0.003 0.022 0.003 0.005 0.003 0.001 0.001 0.000 0.001 0.001 0.001 0.001

WHO2005-TEQPCB 0.145 0.145 0.148 0.148 0.165 0.145 0.147 0.145 0.144 0.111 0.143 0.143 0.144 0.143 0.143

Σdl-PCB 104.622 87.150 187.008 108.548 738.584 97.137 162.652 104.512 49.114 34.591 14.140 32.777 44.387 37.130 32.755

(15)

Table 5: Total marker PCB concentrations obtained by the three different instrumentation methods.

Fish Milk Egg Ground beef

Sample ID

SH farm

SH wild Rainbow Salmon

PAC wild Salmon PAC farm Salmon Atlantic

Dairyland Avalon Fhouse CDN

Harvest

Avalon MHF Regular Lean fresh Supermarket

regular TSQ 1 ul Unit pg/gww pg/glipid PCB6 PCB 28 29.4 0.5 28.5 35.2 137.0 4.4 1.9 17.0 0.4 1.5 2.3 0.0 0.00 0.00 0.00 PCB 52 93.9 63.0 118.9 107.4 451.9 67.6 22.6 87.3 6.0 16.2 22.1 18.3 2.32 2.47 7.20 PCB 101 162.0 99.5 219.1 158.3 915.9 112.3 0.0 130.7 2.8 27.4 30.3 34.9 6.45 8.67 13.07 PCB 138 171.7 114.3 245.2 116.2 989.2 132.9 0.0 234.4 30.8 32.7 16.0 31.5 23.19 25.87 23.70 PCB 153 266.7 195.9 513.1 186.2 1737.1 228.4 179.1 285.3 45.3 33.3 27.8 49.5 35.58 33.11 45.94 PCB 180 86.8 70.2 167.0 28.3 451.3 73.3 0.0 93.0 5.4 0.0 23.2 13.5 9.89 10.64 12.30 ΣPCB6 811 543 1,292 632 4,683 619 204 848 91 111.02 121.70 147.77 77.42 80.75 102.20 TSQ PTV 5 ul Unit pg/g ww pg/g lipid PCB6 PCB 28 27.8 22.2 30.3 38.3 130.0 26.0 36.9 17.0 7.1 20.3 33.5 23.6 0.00 0.00 5.57 PCB 52 90.8 67.8 115.5 106.6 439.1 67.0 161.0 87.3 6.5 33.0 30.5 27.9 0.00 0.00 8.61 PCB 101 160.2 108.6 231.5 159.5 892.8 113.9 57.9 130.7 9.4 42.0 34.2 35.8 0.00 0.00 13.88 PCB 138 159.9 118.8 251.6 113.2 945.7 130.6 122.5 234.4 37.4 30.0 21.9 30.3 0.00 0.00 30.39 PCB 153 275.7 195.6 490.9 176.4 1695.3 227.8 157.3 285.3 49.7 50.6 41.3 63.6 0.00 0.00 46.13 PCB 180 95.1 72.6 188.6 31.5 473.6 78.8 48.0 93.0 15.7 0.0 0.0 11.3 0.00 0.00 16.53 ΣPCB6 810 586 1,309 625 4,576 644 584 848 126 175.93 161.30 192.47 0.00 0.00 121.10 HRMS 2 ul Unit pg/g ww pg/g lipid PCB6 PCB 28 17.0 16.0 25.9 29.5 102.1 17.5 23.0 17.0 7.2 22.6 25.4 17.2 4.59 0.00 8.25 PCB 52 87.3 56.8 117.7 93.9 447.2 58.3 159.5 87.3 4.1 25.9 18.9 19.5 5.97 0.00 4.80 PCB 101 130.7 87.8 209.7 129.0 745.5 91.8 140.0 130.7 0.0 0.0 35.0 39.5 0.00 0.00 5.31 PCB 138 234.4 170.9 415.6 146.1 1392.9 196.0 87.9 234.4 38.0 46.4 30.7 61.2 23.63 33.47 38.91 PCB 153 285.3 205.7 459.4 188.7 1806.5 229.3 150.3 285.3 48.2 28.2 29.8 54.2 31.30 32.34 46.52 PCB 180 93.0 60.6 178.0 34.1 439.1 89.2 66.7 93.0 17.7 0.0 0.0 6.2 12.46 15.50 11.67 ΣPCB6 848 598 1,406 621 4,933 682 627 848 115 123.10 139.75 197.91 77.96 81.31 115.46

(16)

Table 6: Upper and lower bound WHO2005 TEQ for PCDD/PCDFs. Lower bound values were calculated by substituting 0 for all non-detects. Upper bound values had the respective LOQs substituted for non-detects.

Fish Milk Egg Ground beef

SH farm SH wild Rain-bow Salmon PAC wild Salmon PAC farm Salmon Atlantic Dairy-land Avalon Fhouse CDN Harvest

Avalon MHF Regular Lean

fresh

Supermar ket regular

Lower bound / Units pg/g ww pg/g lipid

WHO2005-TEQPCDD 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.035 0.104 0.007 0.000

WHO2005-TEQPCDF 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

WHO2005-TEQPCDD/PCDF 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.035 0.104 0.007 0.000

Upper bound

WHO2005-TEQPCDD 0.299 0.299 0.299 0.299 0.299 0.299 n/a 0.299 0.299 0.299 0.299 0.299 0.377 0.299 0.299

WHO2005-TEQPCDF 0.176 0.176 0.176 0.176 0.176 0.176 n/a 0.176 0.176 0.176 0.176 0.176 0.176 0.176 0.176

(17)

Throughout sample quantification, the different instrument injectors produced slightly different results in regard to the sample TEQ. This was caused due to the quantification of PCB 126 and PCB 169 (Appendix 1). These two congeners have the highest TEFs of 0.1 and 0.03 compared to the other dl-congeners whose values are 0.00003. Values as low as 0.5 pg/g were quantified which significantly changed the results for the lower bound TEQs. Lower bound data was calculated by assuming a value of 0 for all compounds that were not detected. Conversely, upper bound data used the instrumental LOQs in place of all non-detected compounds. This approach is in agreement with the EU regulation for dioxins and PCB in food and feed (EU, 2014). The initial TSQ data was run using a split/splitless injector with a 1 µl injection volume. To obtain qualitative and quantitative results, the instrumental method must be sensitive as well as selective. The column produced very selective data as no interferences were observed between the congeners. Unfortunately, the method lacked the necessary sensitivity. This worked well for the indicator PCB where all congeners could be identified and quantified in all samples but, the concentrations of the dl-PCB were too low to detect. A new PTV method was developed with the goal of by increasing the injection volume from 1 µl to 5 µl in order to detect the low-level dl-PCBs. The resulting data had higher sensitivity thus, more congeners were detected which can be seen in the lower bound TEQs in Table 1. HRMS was used as another method of detection which also complies with EPA 1668C. Data between the different methods were nearly identical for all markers. The variation occurred when looking at the low-level dioxin-like congeners. Due to the high LOQ on the TSQ, data below this limit is suspect to deviation and any positive value found from one of the more toxic congeners (126 and 169) can greatly affect the resulting TEQ.

Quality control samples were run in parallel with the sample batches. A total of three blanks and three LCS were run. Each different instrumental method produced similar result with no

contamination seen in the blanks. All LCS samples processed produced recoveries ranging from 70-130%. A variance between the duplicate samples was not observed. The sample with the highest TEQ was the farmed Pacific spring salmon (as seen in Table 3). This sample gave the highest variation of concentrations when different instrumentation method was used. This was traced back to the quantification of PCB 126 which is the most toxic congener with a TEF of 0.1. Both TSQ runs (1 µl, 5 µl) gave values of 8 pg and 2.2 pg respectively. The HRMS instrument did not detect PCB 126 which led to a significant decrease of the TEQ. This contrasted the wild salmon samples which had concentrations significantly lower than the spring salmon. The results suggest that the limited space/food source of the farmed salmon could have led to the higher concentrations seen, where as the wild salmon could swim in the ocean which allows for the dilution of PCB concentrations.

(18)

Figure 1: Marker PCB values expressed as a percentage of total concentration

Concentrations of the marker PCBs were significantly higher than the dioxin-like PCBs, but were still well below the EU action levels (EU, 2011) with values between 0 - 6.6% (Table 7). No difference was seen in the marker PCB concentrations between the different instrumental methods.

Table 7: EU regulations for maximum levels in foodstuffs.

Beef Egg Milk Fish

Units pg/g fat pg/g fat pg/g fat pg/g wet weight

Σ Marker PCBs 40000 40000 40000 75000

ΣWHO2005-TEQtotal 4 5 5.5 6.5

ΣWHO2005TEQPCDD/PCDF 2.5 2.5 2.5 3.5

Total PCB concentrations for fish ranged from 600 pg/g to 5700 pg/g fresh weight which were by far the highest concentrations in all the samples. This was compared to 7100 pg/g fresh weight found in a Belgian study (Voorspoels, 2008) and 25000 pg/g fresh weight in Finland (Kiviranta, 2004). The differences in these results is most likely due to the location that the samples were taken. For example, the fish present in the Finish study most likely came from the Baltic sea, which is a very shallow body of water in comparison to the Pacific Ocean. With less volume of water to dilute the concentrations of PCBs, higher values would be reasonable to expect from samples taken in that area.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Marker PCBs

PCB 180 PCB 153 PCB 138 PCB 101 PCB 52 PCB 28

(19)

Figure 2: Comparison of dioxin-like PCBs in fish samples

The concentrations in all other samples ranged from 100 pg/g to 250 pg/g lipid. Chicken eggs samples were used for analysis due to their significantly higher lipid content over chicken meat as apposed to other studies where both egg and meat were analysed. This was to avoid using a large sample size.

The two mass production milk samples did not exhibit a large variance between the upper bound TEQs whereas the local “farm fresh” sample had a lower TEQ. The real difference was found when comparing the marker PCBs. Both larger production milks had marker values ranging from 600-850 pg/g lipid where the Farm House contained only 90-125 pg/g lipid. The lower values could be attributed to two factors. The first being the environment that the cows were raised in. This includes different feeds as well as living conditions. These variables could be followed up on given a more time. The next factor could be the extraction of the milk

themselves. DairyLand and Avalon advertised their product to contain 3.5% fat and the Farm House at 4.0%. After extraction, the lipid content found was 1.4 %, 0.8%, and 3.4% respectively. With these results, it is evident that some of the lipid may have been left behind after transferring the organic hexane layer post acid shake. Falsely low lipid content would skew the results higher than then actually are since the results are calculated on a per gram lipid basis.

Sample weights also varied between the different studies. Trout samples that were analyzed in from the great lakes used 15 g wet weight for the analysis (Rawn et al., 2017) whereas only 10 g wet weight was used in the present study. In comparing the results, medium sized trout found in the great lakes had dl-PCB TEQs ranging from 2.9 pg TEQ/g – 16 pg TEQ/g wet weight whereas trout caught in British Columbia has lower values of 0.15 pg TEQ/g – 1.12 pg TEQ/g wet

weight. Overall, PCB concentration were lower in the current study when compared to those that were performed in Europe.

Total PCB concentrations are not useful in estimating toxicity. This is due to the fact that the majority of the concentration of PCBs come from the markers, which are significantly less toxic than the dioxin-like PCBs. The dioxin-like PCBs have their toxic equivalent factors (TEF) calculated based on the similarity to 2,3,7,8-TCDD. The exception is PCB 118 which is both a

0.0 100.0 200.0 300.0 400.0 500.0 PCB 77 PCB 81 PCB 126 PCB 169 PCB 105 PCB 114 PCB 118 PCB 123 PCB 156 PCB 157 PCB 167 PCB 189 pg/ g w et w ei gh t

Dioxin-like PCBs in Fish

Trout Steelhead farmed Trout Stealhead wild

Rainbow trout Salmon - Fresh Street market Sockeye (Wild)

(20)

marker and dioxin-like PCB. Although the present study is not comprehensive, it gives a good indication about the foods in which PCB levels are generally highest

Table 8: Calculation of congener specific PCDD/PCDF LOD and LOQ. LOD is calculated using 3σ. LOQ is 10σ. HRMS (PCDD/PCDF) pg/g LOD pg/g LOQ pg/g pg/g LOD pg/g LOQ pg/g 2378-TCDF 0.010 0.03 0.1 2378-TCDD 0.006 0.018 0.06 12378-PeCDF 0.018 0.054 0.18 12378-PeCDD 0.015 0.045 0.15 23478-PeCDF 0.021 0.063 0.21 123478-HxCDD 0.031 0.093 0.31 123478-HxCDF 0.022 0.066 0.22 123678-HxCDD 0.027 0.081 0.27 123678-HxCDF 0.025 0.075 0.25 123789-HxCDD 0.027 0.081 0.27 123789-HxCDF 0.023 0.069 0.23 1234678-HpCDD 0.038 0.114 0.38 234678-HxCDF 0.023 0.069 0.23 OCDD 0.092 0.276 0.92 1234678-HpCDF 0.017 0.051 0.17 1234789-HpCDF 0.032 0.096 0.32 OCDF 0.058 0.174 0.58

The analysis for PCDD/PCDF gave only 7 quantifiable results (from 17 congeners in 15 samples), five being PCDD and two PCDF. In every case seen, positives were not replicated between the duplicate samples. This furthers the notion that all samples did not contain any significant amounts of dioxins or furans as outliers could have been present. The data points were checked and one of the reasons for variability could be caused by potential hot spots in the tissue. This method has been accredited according to ISO17025 guidelines which validates the procedure used. Since PCDD/PCDF are typically present in these foodstuffs, these results imply that the sample size was too small and not enough fat present for analysis.

4 Conclusion

In comparison to published data around the world, it is quite clear that there is a variation in contamination based on sample location and type of foodstuff (Kiviranta, 2004) (Voorspoels, 2008). For Canada, this first screening using the most eaten foodstuffs and most common sources of food – small markets and large supermarkets - did not result in high concentrations for marker

(21)

PCB or dioxin-like compounds. In comparison to the strictly regulated European market, the concentrations were less than 10% of the EU limits.

The attempts to obtain more quantifiable results by increasing the injection volume or using the more sensitive and selective detector (HRMS) did not really improve the number of quantifiable results. In the absence of Canadian regulation as to the sampling and limit values for foods, the results cannot be put into perspective for the Canadian or North American market and cannot be compared with the EU situation. At least, one study is needed to use the recommended EU approach by analysing samples with at least 10g of fat present in the sample to make a strong conclusion that the relatively scarcely populated West Canadian region provides a “natural” dilution of PCB and dioxin-like compounds and concentrations below those found in the more densely populated European region where often the local separation between industry ab agriculture is not possible.

5 References

Adamse, P., Schoss, S., Theelen, R. and Hoogenboom, R. (2016). Levels of dioxins and dioxin-like PCBs in food of animal origin in the Netherlands during the period 2001–2011. Food

Additives & Contaminants: Part A, 34(1), pp.78-92.

Bemis, J., Nazarenko, D. and Gasiewicz, T. (2005). Coplanar Polychlorinated Biphenyls

Activate the Aryl Hydrocarbon Receptor in Developing Tissues of Two TCDD-Responsive lacZ Mouse Lines. Toxicological Sciences, 87(2), pp.529-536.

Bernard, A., Hermans, C., Broeckaert, F., De Poorter, G., De Cock, A. and Houins, G. (1999). Food contamination by PCBs and dioxins. Nature, 401(6750), pp.231-232.

Bowes, G. and Jonkel, C. (1975). Presence and Distribution of Polychlorinated Biphenyls (PCB) in Arctic and Subarctic Marine Food Chains. Journal of the Fisheries Research Board of

Canada, 32(11), pp.2111-2123.

Canadian Environmental Protection Act (1999). PCB Regulations SOR/2008-273, http://laws-lois.justice.gc.ca/eng/regulations/SOR-2008-273/page-1.html

Darnerud, P., Atuma, S., Aune, M., Bjerselius, R., Glynn, A., Grawé, K. and Becker, W. (2006). Dietary intake estimations of organohalogen contaminants (dioxins, PCB, PBDE and chlorinated pesticides, e.g. DDT) based on Swedish market basket data. Food and Chemical Toxicology, 44(9), pp.1597-1606.

Drouillard, K., Jezdic, I., O’Rourke, S., Gewurtz, S., Raeside, A., Leadley, T., Drca, P. and Douglas Haffner, G. (2013). Spatial and temporal variability of PCBs in Detroit River water assessed using a long-term biomonitoring program. Chemosphere, 90(1), pp.95-102.

Erickson, B. (1999). Focus: Dioxin food crisis in Belgium. Analytical Chemistry, 71(15), pp.541A-543A.

(22)

EU (2014). COMMISSION REGULATION (EU) No 589/2014 of 2 June 2014 laying down methods of sampling and analysis for the control of levels of dioxins, dioxin-like PCBs and non-dioxin-like PCBs in certain foodstuffs and repealing Regulation (EU) No 252/2012. Official Journal of the European Union L164/18-L164/40

EU (2011). COMMISSION REGULATION (EU) No 1259/2011 of 2 December 2011 amending Regulation (EC) No 1881/2006 as regards maximum levels for dioxins, dioxin-like PCBs and non dioxin-like PCBs in foodstuffs. Official Journal of the European Union L320/18-L320/23 Fitzgerald, E., Brix, K., Deres, D., Hwang, S., Bush, B., Lambert, G. and Tarbell, A. (1996). Polychlorinated Biphenyl (Pcb) and Dichlorodiphenyl Dichloroethylene (Dde) Exposure Among Native American Men from Contaminated Great Lakes Fish and Wildlife. Toxicology and

Industrial Health, 12(3-4), pp.361-368.

Fiedler, H., Cooper, K., Bergek, S., Hjelt, M. and Rappe, C. (1997). Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (P)CDD/PCDF) in food samples collected in Southern Mississippi, USA. Chemosphere, 34(5-7), pp.1411-1419.

Focant, J., Pirard, C. and De Pauw, E. (2004). Levels of PCDDs, PCDFs and PCBs in Belgian and international fast food samples. Chemosphere, 54(1), pp.137-142.

Holoubek, I. (2001). Polychlorinated Biphenyl (PCB) Contaminated Sites Worldwide. In Needham L., Fiedler H., Robertson L., & Hansen L. (Eds.), PCBs: Recent Advances in

Environmental Toxicology and Health Effects (pp. 17-26). University Press of Kentucky.

Retrieved from http://www.jstor.org/stable/j.ctt130j2pw.7

Huwe, J. and Larsen, G. (2005). Polychlorinated Dioxins, Furans, and Biphenyls, and Polybrominated Diphenyl Ethers in a U.S. Meat Market Basket and Estimates of Dietary Intake. Environmental Science & Technology, 39(15), pp.5606-5611.

Huwe, J., Pagan-Rodriguez, D., Abdelmajid, N., Clinch, N., Gordon, D., Holterman, J., Zaki, E., Lorentzsen, M. and Dearfield, K. (2009). Survey of Polychlorinated Dibenzo-p-dioxins,

Polychlorinated Dibenzofurans, and Non-ortho-polychlorinated Biphenyls in U.S. Meat and Poultry, 2007−2008: Effect of New Toxic Equivalency Factors on Toxic Equivalency Levels, Patterns, and Temporal Trends. Journal of Agricultural and Food Chemistry, 57(23), pp.11194-11200.

IARC (2016). Polychlorinated Biphenyls and Polybrominated Biphenyls, Vol 107 (Lyon, France: IARC).

Khodabocus, F. and Balgobin, K. (2011). Implementation and Practical Benefits of ISO/IEC 17025:2005 in a Testing Laboratory. University of Mauritius Research Journal, 17(1).

Kiviranta, H., Ovaskainen, M. and Vartianinen, T. (2004). Market basket study on dietary intake of PCDD/Fs, PCBs, and PBDEs in Finland. Environmental International, 30(7), 99. 923-932. Lauby-Secretan, B., Loomis, D., Baan, R., El Ghissassi, F., Bouvard, V., Benbrahim-Tallaa, L., Guha, N., Grosse, Y. and Straif, K. (2015). Use of mechanistic data in the IARC evaluations of

(23)

the carcinogenicity of polychlorinated biphenyls and related compounds. Environmental Science

and Pollution Research, 23(3),.2220-2229.

Newsome, W., Davies, D. and Sun, W. (1998). Residues of polychlorinated biphenyls (PCB) in fatty foods of the Canadian diet. Food Additives and Contaminants, 15(1), pp.19-29.

Rawn, D., Dowd, M., Scuby, M., Pantazopoulos, P. and Feeley, M. (2017). Polychlorinated Biphenyls and Polychlorinated Dioxins–Furans in Lake Trout and Whitefish Composite Samples from Commercial Fisheries in Lakes Erie, Huron, and Superior. Journal of Food Protection, 80(8), pp.1228-1238.

Schecter, A. and Li, L. (1997). Dioxins, dibenzofurans, dioxin-like PCBs, and DDE in U.S. fast food, 1995. Chemosphere, 34(5-7), pp.1449-1457.

Scott, L., Staskal, D., Williams, E., Luksemburg, W., Urban, J., Nguyen, L., Haws, L., Birnbaum, L., Paustenbach, D. and Harris, M. (2009). Levels of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in southern Mississippi catfish and estimation of potential health risks. Chemosphere, 74(7), pp.1002-1010.

UNEP (2001). Stockholm Convention on Persistent Organic Pollutants. United Nations Environment Programme, www.pops.int

US-EPA (2010): Method 1668 Revision C, “Method 1668, Revision C: Chlorinated Biphenyl Congeners in Water, Soil, Sediment, and Tissue by HRGC/HRMS”. US Environmental Protection Agency, Office of Water, April 2010

Van den Berg, M., Birnbaum, L.S., Denison, M., De Vito, M., Farland, W., Feeley, M., Fiedler, H., Hakansson, H., Hanberg, A., and Haws, L. (2006). The 2005 World Health Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds. Toxicological sciences 93, 223-241

Voorspoels, S., Covanci, A. and Neels, H. (2008). Dietary PCB intake in Belgium.

Figur

Updating...

Relaterade ämnen :